High performance semiconductor devices, especially those with p-n junctions, are historically made with single crystal semiconductor materials. The use of such single crystal materials for semiconductor devices effectively avoids charged carrier scattering and minority carrier recombination when compared to using non-single crystal semiconductor materials that have grain boundaries. The scattering of charge carriers adversely reduces the drift mobility and the diffusion of the charged carriers, which leads to degraded performance of devices, such as photonic devices. Degraded performance is manifested by increased resistance and a decrease in optical to electrical conversion efficiency, for example. Even when different semiconductor materials were employed together in a semiconductor device, such as in a heterostructure or heterojunction device, single crystal semiconductor materials are generally chosen based on their respective lattice structures. This is to insure that the realized structure is effectively a single crystal structure as a whole. Similarly, nanostructures including, but not limited to, nanowires and nanodots are typically nucleated and grown from single crystal substrates, in part to capitalize on the uniform nature of the lattice of such substrates that provides required crystallographic information for the nanostructures to be aligned.
Photovoltaic cells are one type of photonic device that are the subject of much interest due to high energy costs and U.S. dependence on fossil fuel from foreign sources. Photodetectors are another type of photonic device of particular interest. The efficiency and quality of photovoltaic cells have improved significantly over the last 10 years. Efforts to lower the cost of photovoltaic cells have been directed at alternative materials and manufacturing methods.
Amorphous and other non-single crystal semiconductor materials, such as polycrystalline semiconductor materials, have attracted attention at least for potential cost savings in photovoltaic applications. While having the disadvantage of lower efficiencies that is associated with multiple grain boundaries in the non-single crystal semiconductor materials, such materials can be considerably cheaper to manufacture than their single crystal counterparts. There are applications where the lower cost of producing the semiconductor device from non-single crystal materials may outweigh any loss of performance that may result. However, this trade-off is simply not an option for many photonic applications.